Concept of Measurement The ability to demonstrate change or relationship and to communicate those changes to others. Describes the quality or quantity of an existing variable e.g. ROM and strength
Concept of Measurement Evaluate a patients condition and response to treatment. We measure change Is the process of assigning numerals to objects to represent quantitative characteristics according to certain rules.
Variable: a characteristic that can be manipulated or observed and that can take on different values, either Quantitatively Qualitatively The ability to measure a variable is dependent on one’s ability to define it.
Continuous Variable: can take on any value along a continuum within a defined range (e.g. ROM 50.5 0 or 50 0) Discrete Variable: described only in whole numbers (e.g. HR) Construct: an abstract concept that is invented to represent unmeasureable behaviors or ideas
Rules of Measurement Nominal Ordinal Interval Ratio
Rules of Measurement Nominal - numerals represent category labels only, classifications Sex Blood Type Clinical Diagnosis
Ordinal - numbers indicate rank order of observations MMT Pain Functional Status
Interval - equal intervals between numbers but not related to true zero, therefore not representing an absolute quantity IQ F0 Calendar Years
Ratio - numbers represent units with equal intervals, measured from true zero Weight Strength BP
Reliability Reliability - the extent to which a measurement is consistent and free from error ( repeatability) Usefulness of measurement in clinical research and decision making depends on the extent to which the therapist can rely on data as accurate and meaningful indicator of behavior or attribute
Reliability Usefulness of measurement in clinical research and decision making Depends on the extent to which the therapist can rely on data as accurate and meaningful indicator of behavior or attribute
Measurement Errors X = T + E Observed Score True Score Error Measurements are rarely perfectly reliable Error in Measurement X = T + E Observed Score True Score Error Errors can be- Systematic Errors: are predictable errors of measurement Random Errors: due to chance (unpredictable)
Source of Measurement Error Rater error Inaccuracies in the measuring instruments variability of the characteristics being measured
Source of Measurement Error Development of testing instruments involve a specific protocol that maximizes the reliability of the instrument. Errors are identified and then controlled or eliminated: Careful Planning Clear Operational Definitions Inspection of Equipment
Estimate of Reliability True Score Variance True Score Variance and Error Variance T T + E T + E T reliability increases 1.00 perfect reliability .00 no reliability
Correlation: degree of association between two sets of data or variables. Correlation between height and weight Not cause and effect research State two variables are related (X,Y) No true variable manipulation Acceptable levels of positive and negative correlations A level of significance 1.0 perfect .75 good .50 poor
Test - retest reliability Establishes that an instrument is capable of measuring a variable with consistency. Analysis called test-rest reliability coefficient. Effected by: Testing Effects - practice or carry over effect test Rater Bias - same rater can be influenced by the memory of the first score. Controlled by blinding tester.
3-Test-Retest Interval time 4-Carryover & testing effects Rater reliability - Rater is part of measurement system, in some cases is the actual instrument.
Intrarater Reliability - refers to the stability of data recorded by one individual across two or more trials Interrater Reliability - concerns variation between two or more raters who measure the same group of subjects
Validity of Measurement Validity - concerns the extent to which an instrument measures what is is intended to measure. Implies that measurement is relatively free from error. Valid test must also be reliable. Low reliability low validity High reliability is not automatically valid Validity helps us make inferences about variables on relevant observable behavior or responses. Inferences go beyond the simple value assigned to them.
Specificity of Validity Face Validity - instrument appears to test what it is supposed to Content Validity - indicates that the items that make up an instrument adequately sample the universe of content that defines that variable being measured
Specificity of Validity Criterion-related validity - indicates the outcomes of one instrument, the target test that can be used as a substitute measure for an established gold standard criterion test, can be concurrent or predictive
Specificity of Validity Concurrent validity - establishes validity when two measures are taken at relatively the same time. Predictive validity - establishes that the outcome of the target test can be used to predict a future criterion score. Prescriptive validity-establishes that the interpretation of a measurement is appropriate for determining effective intervention
Specificity of Validity Construct validity - establishes the ability of an instrument to measure an abstract construct and the degree to which the instrument reflects the theoretical components of the construct.
Evaluating Diagnostic Procedures A diagnostic test is used to screen for the presence or absence of a disease or abnormal condition Dichotomous Categorical Continuous
Evaluating Diagnostic Procedures Sensitivity-the ability of the test to obtain a positive test when the target condition is present
Evaluating Diagnostic Procedures Specificity-Is the ability of the test to obtain a negative test when the condition is really absent
Measuring Change Goal of tx is to effect a positive change Difference between outcome and initial is change or gain score Used to analyze the effect of tx or intervention
Goals Measure change in an individual performance or condition Measure differences between individuals in the amount of change Seek to identify factors that contribute to a good response Intend to draw inferences about tx effects by looking at group differences
Validity of change scores Level of measurement Reliability Stability Linearity